BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20181221T160730Z
LOCATION:C146
DTSTART;TZID=America/Chicago:20181113T160000
DTEND;TZID=America/Chicago:20181113T163000
UID:submissions.supercomputing.org_SC18_sess218_pap488@linklings.com
SUMMARY:Lessons Learned from Analyzing Dynamic Promotion for User-Level Th
 reading
DESCRIPTION:Paper\nData Analytics, Performance, Programming Systems, Stora
 ge, Tools, Visualization, Tech Program Reg Pass\n\nLessons Learned from An
 alyzing Dynamic Promotion for User-Level Threading\n\nIwasaki, Amer, Taura
 , Balaji\n\nA performance vs. practicality trade-off exists between user-l
 evel threading techniques. The community has settled mostly on a black-and
 -white perspective; fully fledged threads assume that suspension is immine
 nt and incur overheads when suspension does not take place, and run-to-com
 pletion threads are more lightweight but less practical since they cannot 
 suspend. Gray areas exist, however, whereby threads can start with minimal
  capabilities and then can be dynamically promoted to acquire additional c
 apabilities when needed. This paper investigates the full spectrum of thre
 ading techniques from a performance vs. practicality trade-off perspective
  on modern multicore and many-core systems. Our results indicate that achi
 eving the best trade-off highly depends on the suspension likelihood; dyna
 mic promotion is more appropriate when suspension is unlikely and represen
 ts a solid replacement for run to completion, thanks to its lower programm
 ing constraints, while fully fledged threads remain the technique of choic
 e when suspension likelihood is high.
URL:https://sc18.supercomputing.org/presentation/?id=pap488&sess=sess218
END:VEVENT
END:VCALENDAR

